Theory
The steady solvers couple pressure and velocity with the SIMPLE algorithm (Patankar and Spalding, 1972): each iteration solves the momentum equation with the current pressure, then a pressure equation that makes the velocity field conserve mass. The pressure correction of plain SIMPLE neglects the velocity corrections of the neighbouring cells, so it overshoots; only a fraction of it is kept, equation (1). SIMPLEC (Van Doormaal and Raithby, 1984) keeps a consistent form of those neighbour terms (OpenFOAM: consistent yes) and needs no pressure relaxation, so .
The pressure is relaxed as a FIELD (OpenFOAM relaxationFactors.fields), whereas for velocity and the turbulence quantities the EQUATIONS are relaxed (see Velocity relaxation).
On auto all relaxation factors move together with the solver settings:
| solver setting | p | U (start-up / final) | k, ω | ρ | e |
|---|---|---|---|---|---|
| incompressible, SIMPLEC (default) | 1.0 | 0.7 / 0.9 | 0.7 | - | - |
| incompressible, plain SIMPLE | 0.3 | 0.7 / 0.7 | 0.7 | - | - |
| compressible, SIMPLE (auto) | 0.3 | 0.7 / 0.7 | 0.7 | 0.1 | 0.7 |
| compressible, transonic SIMPLE | 1.0 | 0.9 / 0.9 | 0.9 | 1.0 | 0.8 |
| compressible, pseudo-transient | not relaxed (all factors 1) |
- For the compressible solver, auto uses plain SIMPLE (
consistent no) whatever the SIMPLEC switch says. - A typed pressure relaxation makes the SIMPLEC switch take effect as set: with SIMPLEC on, the other auto factors then take the SIMPLEC values (U 0.7 / 0.9, k and ω 0.7, ρ 1.0, e 0.9 with heat transfer, else 0.7).
- If the start-up phase diverges, the run restarts once from the initial fields with the conservative set p 0.3, U 0.7, k 0.7, ω 0.7, ρ 0.1 and
consistent no. The pseudo-transient solver halves its Courant number instead.
Equations
| where | meaning | unit |
|---|---|---|
| Pressure of the previous iteration | Pa (m²/s² incompressible) | |
| Pressure after this iteration's pressure equation | Pa (m²/s² incompressible) | |
| Pressure relaxation factor | - | |
| Velocity relaxation factor (final phase) | - |
Default and rule
- Default
- auto: 1.0 (SIMPLEC), 0.3 (plain SIMPLE and every compressible run except transonic), 1.0 (transonic SIMPLE)Computed by the programMeasured
- Rule
Auto (see the table). A typed value is used as typed in both phases. It gets a warning when it does not suit the solver: below 0.9 with SIMPLEC (SIMPLEC's headroom is not used), with plain SIMPLE, and any typed value with SIMPLEC on the compressible solver.
- Source
Validation on the two-inlet test muffler, 600 iterations on the same mesh; pressure initial residual at the end (p/U/k/ω):
- plain SIMPLE 0.3/0.7/0.7/0.7: 7.1e-2
- plain SIMPLE 0.3/0.5/0.5/0.5: 2.3e-2
- plain SIMPLE 0.3/0.7/0.3/0.3: 5.3e-2
- plain SIMPLE 0.3/0.5/0.3/0.3: 4.2e-2
- SIMPLEC 1/0.9/0.9/0.9: 1.1e-2, with the steadiest pressure drop (0.5 % scatter)
- SIMPLEC 1/0.7/0.7/0.7: 1.6e-2
The pressure drop stayed between 1918 and 1951 Pa in every case. SIMPLEC is the default on this basis.
Compressible solver, case V1 (a 2.0 m multi-chamber muffler with a perforated pipe and three absorptive packings, 691k cells): with SIMPLEC, p 1.0 and ρ 1.0, the compressible solver aborted within 2-8 iterations at 0.5 and at 2.5 kg/s, also from a field that plain SIMPLE had already settled; plain SIMPLE (p 0.3, ρ 0.1) ran. This is the basis of the compressible auto set. The transonic set is that of OpenFOAM's internal hot-gas tutorial (the compressible solver/squareBend). The sum rule for SIMPLE is a common practical guideline; the 1.1 warning threshold is the program's choice.
Valid range 0 < α ≤ 1
Values
. Smaller values are more robust and converge more slowly.
SIMPLEC on the compressible solver
A typed pressure relaxation applies SIMPLEC to the compressible solver as switched; with it aborted on a real muffler.
Pseudo-transient solver
the transient compressible solver with local time stepping relaxes nothing (all factors 1); its stability comes from the local Courant number (Local Courant number (pseudo-transient)).
Related 7
References 2
- Patankar, S. V. and Spalding, D. B. (1972). A calculation procedure for heat, mass and momentum transfer in three-dimensional parabolic flows. International Journal of Heat and Mass Transfer, 15(10), 1787-1806. DOI 10.1016/0017-9310(72)90054-3
- Van Doormaal, J. P. and Raithby, G. D. (1984). Enhancements of the SIMPLE method for predicting incompressible fluid flows. Numerical Heat Transfer, 7(2), 147-163. DOI 10.1080/01495728408961817